COMPLEXIFICATION KEITH CONRAD 1. Introduction We want to describe a procedure for enlarging real vector spaces to complex vector spaces n in a natural way. For instance, the natural complex analogues of R ,Mn(R), and R[X] n are C ,Mn(C) and C[X]. Why do we want to complexify real vector spaces? One reason is related to solving equations. If we want to prove theorems about real solutions to a system of real linear equations or a system of real linear differential equations, it can be convenient as a first step to examine the complex solution space. Then we would try to use our knowledge of the complex solution space (for instance, its dimension) to get information about the real solution space. In the other direction, we may want to know if a subspace of Cn which is given to us as the solution set to a system of complex linear equations is also the solution set to a system of real linear equations. We will find a nice way to describe such subspaces once we understand the different ways that a complex vector space can occur as the complexification of a real subspace. We will give two descriptions of the complexification process, first in terms of a two- fold direct sum (Section2) and then in terms of tensor products (Section3). The tensor product viewpoint is the more far-reaching one, but seeing how the direct sum method of complexification expresses everything may help convince the reader that there is nothing unexpected about this use tensor products. Moreover, idiosyncratic aspects of the direct sum construction will turn out to be completely natural features of the tensor product construction. After comparing the two constructions, we will see (Section4) how to use a special structure on a complex vector space, called a conjugation, to describe the real subspaces which have the given complex vector space as its complexification. References on this topic are [1, pp. 79{81], [2, x77], and [3, pp. 262{263]. In [1] and [3] tensor products are used, while [2] uses direct sums. 2. Complexifying with Direct Sums Let W be a real vector space. To create from W a naturally associated complex vector space, we need to give a meaning to (a + bi)w, where a + bi 2 C and w 2 W . Whatever it might mean, we'd like to have (a + bi)w = aw + biw = aw + ibw; and since there is no given meaning to ibw (that is, multiplying elements of W by i has no definition since W is a real vector space), aw + ibw should be thought of as a formal sum of aw and bw with the i factor keeping them apart. A legitimate way to make sense of this is to treat aw + ibw as the ordered pair (aw; bw), and that leads to the following definition. 1 2 KEITH CONRAD Definition 2.1. The complexification of a real vector space W is defined to be WC = W ⊕ W , with multiplication law (a + bi)(w1; w2) = (aw1 − bw2; bw1 + aw2), where a and b are real. This rule of multiplication is reasonable if you think about a pair (w1; w2) in W ⊕ W as a formal sum w1 + iw2: (a + bi)(w1 + iw2) = aw1 + aiw2 + biw1 − bw2 = (aw1 − bw2) + i(bw1 + aw2): In particular, (2.1) i(w1; w2) = (−w2; w1): It is left to the reader to check that WC is a complex vector space by the multiplication 0 0 rule in Definition 2.1( e.g., z(z (w1; w2)) = zz (w1; w2)). Since i(w; 0) = (0; w); we have (2.2) (w1; w2) = (w1; 0) + (0; w2) = (w1; 0) + i(w2; 0): Thus elements of WC formally look like w1 + iw2, except iw2 has no meaning while i(w2; 0) does: it is (0; w2). The real subspaces W ⊕ f0g and f0g ⊕ W of WC both behave like W , since addition is componentwise and a(w; 0) = (aw; 0) and a(0; w) = (0; aw) when a is real. The R-linear function w 7! (w; 0) will be called the standard embedding of W into WC. So we treat W ⊕ f0g as the “official” copy of W inside WC and with this identification made we can regard WC as W + iW using (2.2). Example 2.2. Taking W = R, its complexification RC is the set of ordered pairs of real numbers (x; y) with (a + bi)(x; y) = (ax − by; bx + ay). Since (a + bi)(x + yi) = (ax − by) + (bx + ay)i, RC is isomorphic to C as C-vector spaces by (x; y) 7! x + yi. n n Example 2.3. The complexifications of R ,Mn(R), and R[X] are isomorphic to C , n n Mn(C), and C[X], by sending an ordered pair (w1; w2) in R ⊕ R ,Mn(R) ⊕ Mn(R), or n R[X] ⊕ R[X] to w1 + iw2 in C ,Mn(C), or C[X]. 2 2 For instance, we can identify the complexification (R )C with C (as complex vector 2 2 2 1 spaces) by (w1; w2) 7! w1 + iw2 2 R + iR = C , and this sends the basis vectors 0 and 0 2 2 1 0 2 2 1 of R ⊂ (R )C to 0 and 1 in C , which are the standard basis vectors of C as a n complex vector space. More generally, the identifications of (R )C; Mn(R)C, and R[X]C n n with C ,Mn(C), and C[X] turn every real basis of R ,Mn(R), and R[X] (viewed inside n their complexifications by the standard embedding) into a complex basis of C ,Mn(C), and C[X]. Theorem 2.4. If W = 0 then WC = 0. If W 6= 0 and fejg is an R-basis of W then f(ej; 0)g is a C-basis of WC. In particular, dimC(WC) = dimR(W ) for all W . Proof. That W being zero implies WC is zero is easy. Now take W 6= 0 with basis fejg. For (w1; w2) 2 WC, writing w1 and w2 as R-linear combinations of the ej's shows every element of WC is an R-linear combination of the (ej; 0)'s and (0; ej)'s. Since (0; ej) = i(ej; 0), using C-linear combinations we can write every element of WC in terms of the vectors (ej; 0). Therefore f(ej; 0)g is a C-linear spanning set of WC. To show it is linearly COMPLEXIFICATION 3 independent over C (and thus is a basis), suppose we can write (0; 0) as a (finite!) C-linear combination of the vectors (ej; 0), say (a1 + ib1)(e1; 0) + ··· + (am + ibm)(em; 0) = (0; 0) for some real aj and bj. This is the same as (a1e1 + ··· + amem; b1e1 + ··· + bmem) = (0; 0): P P Therefore ajej = 0 and bjej = 0 in W . From linear independence of the ej's over R, all the coefficients aj and bj are 0, so aj + ibj = 0 for all j. Example 2.5. Treating C as a real vector space, its complexification is not C. Indeed CC has to have complex dimension 2 since C has real dimension 2 (Theorem 2.4). Explicitly, CC = C ⊕ C with (a + bi)(z1; z2) = (az1 − bz2; bz1 + az2). More generally, if W is a complex vector space the complexification WC does not know about the original complex scaling on W ; the construction of WC only uses the real vector space structure of W and all information in advance about being able to multiply by i on W is gone. A real m × n matrix, as an R-linear transformation Rn ! Rm, can be viewed in a natural way as a function Cn ! Cm and it becomes a C-linear transformation. The next two theorems show how this process looks from the viewpoint of complexifications. Theorem 2.6. Every R-linear transformation ': W ! W 0 of real vector spaces extends in a unique way to a C-linear transformation of the complexifications: there is a unique 0 C-linear map 'C : WC ! WC making the diagram ' W / W 0 0 WC / W 'C C commute, where the vertical maps are the standard embeddings of real vector spaces into their complexifications. Proof. If such a C-linear map 'C exists, then the commutativity of the diagram says 'C(w; 0) = ('(w); 0) for all w 2 W . Therefore when (w1; w2) is an element of WC, 'C(w1; w2) = 'C(w1; 0) + 'C(0; w2) = 'C(w1; 0) + 'C(i(w2; 0)) = 'C(w1; 0) + i'C(w2; 0) = ('(w1); 0) + i('(w2); 0) = ('(w1); 0) + (0;'(w2)) = ('(w1);'(w2)): This tells us what 'C must be (notice the step where we used C-linearity). So now we just turn around and define 'C : WC ! WC by 'C(w1; w2) := ('(w1);'(w2)): We need to check 'C is C-linear. Since 'C is clearly R-linear, the only thing to check is that 'C commutes with multiplication by i. This is left to the reader, using (2.1). We call 'C the complexification of '. 4 KEITH CONRAD Theorem 2.7. Let W and W 0 be nonzero finite-dimensional real vector spaces. For an R- 0 0 0 linear transformation ': W ! W and R-bases feig and fejg of W and W , the matrix for 0 ' with respect to these bases is the matrix for 'C : WC ! WC with respect to the C-bases 0 f(ei; 0)g and f(ej; 0)g of the complexifications.